Abstract:Neurally inspired robotics already has a long history that includes reactive systems emulating reflexes, neural oscillators to generate movement patterns, and neural networks as trainable filters for high-dimensional sensory information. Neural inspiration has been less successful at the level of cognition. Decision-making, planning, building and using memories, for instance, are more often addressed in terms of computational algorithms than through neural process models. To move neural process models beyond reactive behavior toward cognition, the capacity to autonomously generate sequences of processing steps is critical. We review a potential solution to this problem that is based on strongly recurrent neural networks described as neural dynamic systems. Their stable states perform elementary motor or cognitive functions while coupled to sensory inputs. The state of the neural dynamics transitions to a new motor or cognitive function when a previously stable neural state becomes unstable. Only when a neural robotic system is capable of acting autonomously does it become a useful to a human user. We demonstrate how a neural dynamic architecture that supports autonomous sequence generation can engage in such interaction. A human user presents colored objects to the robot in a particular order, thus defining a serial order of color concepts. The user then exposes the system to a visual scene that contains the colored objects in a new spatial arrangement. The robot autonomously builds a scene representation by sequentially bringing objects into the attentional foreground. Scene memory updates if the scene changes. The robot performs visual search and then reaches for the objects in the instructed serial order. In doing so, the robot generalizes across time and space, is capable of waiting when an element is missing, and updates its action plans online when the scene changes. The entire flow of behavior emerges from a time-continuous neural dynamics without any controlling or supervisory algorithm.

Motor primitive and sequence self-organization in a hierarchical recurrent neural network

Hierarchical Dynamical Models of Motor Function

Towards a self-organizing pre-symbolic neural model representing sensorimotor primitives

Autonomous Sequence Generation for a Neural Dynamic Robot: Scene Perception, Serial Order, and Object-Oriented Movement

Self-Organization Of Spiking Neural Network Generating Autonomous Behavior In A Miniature Mobile Robot

Learning Movement Sequences with a Delayed Reward Signal in a Hierarchical Model of Motor Function.

Self-organized adaptation of a simple neural circuit enables complex robot behaviour

Hierarchical generative modelling for autonomous robots

Self-Organization of Spiking Neural Network Generating Autonomous Behavior in a Real Mobile Robot

Hierarchical control of goal-directed action in the cortical–basal ganglia network

Hierarchical behavior control by a single class of interneurons

Self-organizing Continuous Attractor Networks and Motor Function

Hierarchical Self-Programming in Recurrent Neural Networks

Hierarchical motor control in mammals and machines

Embodied robots driven by self-organized environmental feedback

Learning and Encoding Motor Primitives for Limb Actions in a Brain-Like Computation Approach.

Neural probabilistic motor primitives for humanoid control

Recursion in action: An fMRI study on the generation of new hierarchical levels in motor sequences

Self-organization of multi-layer spiking neural networks

Evolving Network Motifs Based Morphogenetic Approach To Self-Organizing Robotic Swarms

A robot model of the basal ganglia: Behavior and intrinsic processing